Note that any discussion of the prior art throughout the specification should in no way be considered as an admission that such prior art is widely known or forms part of common general knowledge in the field.
All positive displacement type compressors suffer to a greater or lesser extent from possessing a higher than desirable ‘dead volume’ (hereafter DV). The DV is the volume remaining in the working chamber after the piston has reached the TDC position. Ideally, that volume would generally be zero. The outcome of it not being zero is that the compressed gas remaining in the DV is then not forced out through the exit valve into a receiving vessel, but is re-expanded by movement of the piston and is returned to the next intake stroke. As a result the volumetric efficiency of the compressing machine is greatly impaired. Therefore to then achieve the desired quantity of delivered compressed gas requires that the machine has to possess a larger swept volume. A larger machine implies increased weight, bulk and manufacturing cost as well as increased mechanical friction and other energy losses.
A range of potential Wankel-type compressor concepts exist which incorporate epitrochoidal type housings. The most promising are the type with a three cornered rotor rotating inside a two lobed epitrochoidal housing; and the similarly principled type with a two cornered rotor inside a one lobed housing. The former (conventionally designated the 2:3 type, the latter being the 1:2 type) has been built by several manufacturers as an IC engine in considerable volume. However, when first proposed some 60 years ago, both types were equally put forward as potential gas compressors.
The main reason that the 2:3 type has failed to be successful in the market place for the compressor application is related to the DV problem. In a practical current-art design, the DV is typically 10 to 16% which is too high for an efficient machine. It is true that if a higher R/e value is selected, (“R” being the radius of the rotor and “e” being the eccentricity of the shaft on which the rotor is mounted), then a somewhat lower DV can be achieved. But a greater R/e results in a bigger and heavier machine with higher mechanical friction.
The alternative 1:2 type can achieve a DV significantly lower than the 2:3 type, particularly if a higher R/e value is selected. Therefore considerably more efforts have been made in the past to develop such a 1:2 compressor. However, when utilising such a high R/e value, this 1:2 machine then suffers from possessing a very small diameter stationary gear and drive shaft with considerably less than the ideal torque capability. If any significant dynamic torque loading were then to occur, due to dynamic torsional vibration acting on the input drive shaft as may be caused by the inherent and known torque reversal problem for example (as discussed in U.S. Pat. No. 4,218,199A), the gear or shaft may be overstressed and fail. Hence this type has not proved suitable for general industrial usage.
Some features of a potential design of epitrochoidal type compressor are presented in patent application no. GB2215403 (The Hydrovane Compressor Company Limited). This document also lists the many problems, particularly relating to high friction losses, associated with the sliding vane type of positive displacement compressor. These problems lead to low energy efficiency, particularly when operating at higher speeds or part load or producing pressure greater than 4 bar or so. Nevertheless many manufacturers currently supply large numbers of this type to the market place, despite the need for higher energy efficiency having become an increasingly important consideration.
In an attempt to provide a compressor with increased efficiency relative to the sliding vane type GB2215403 identified the rotary type with epitrochoidal housing as a promising candidate, particularly with regard to its superior gas sealing principles, mechanical efficiency, and part-load control characteristics. GB2215403 identifies the need to seal the HP chamber from the LP chamber around the TDC position; and proposed to use stationary seal pieces located in the inner surface of the housing circumferentially positioned at the minor axis of the housing which engage with the flank surface of the rotor to achieve this end.
However when the chamber positioned in the vicinity of the TDC position is divided into two sectors by the presence of such sealing means, the gas pressure now acting on the two areas of the rotor flanks on either side of the seal is very disparate. This results in a high torque being applied to the rotor which then imposes a high load on some teeth of the rotor and stationary gears in a repetitive and cyclic manner. Therefore, unless special design considerations are applied, these gears would probably suffer fatigue failure if the machine was used to produce gas pressure in the frequently required range of 5 to 8 bar or higher. GB2215403 failed to identify or hence address this important issue; and the design is therefore deficient.
By utilising stationary seal pieces located at each minor axis in the housing, the design of GB2215403 is unable to utilise the conventional arrangement of apex seals located at the apices of the rotor, such rotating seals being generally incompatible with the design to use a stationary seal located in the housing, as each moving apex seal piece would impact with each stationary seal piece once for each revolution of the rotor and inevitably cause damage.
Hence, to avoid this second deficiency, GB2215403 proposed not to use apex seal pieces located at the rotor apices, but to rely on the necessary gas sealing at these places being achieved by designing and manufacturing the rotor to provide a very small radial working clearance of 0.1 mm maximum between the rotor outer periphery at the apices and the epitrochoidal inside surface of the housing for all positions of the rotor.
However, design and manufacturing experience with the Wankel engine indicates that it is not practical or economic to specify such a small clearance between the rotor and the bore of the housing because many tolerances are involved in the manufacture of the major related components, such as rotor with internal gear, stationary gear, eccentric shaft, end plates, and rotor housing, etc., which may each contribute additively to the required working clearance between the rotor and the housing bore.
A major contributor to this need for clearance is the necessary or inevitable backlash between the rotor and stationary gears, as well as the angular and radial location accuracy of each of these gears in their respective components. When, during rotation, the rotor apices are situated at the minor or major axis of the housing, the backlash plus gear angular location tolerances do not materially influence the radial clearance value between rotor apices and housing bore; but when the rotor apices are in between these positions the rotational “free play” of the rotor, combined with the many potential radial location errors, may allow the apices to collide with the housing surface unless a positive clearance always exists. If this mechanical contact were to occur, the machine may fail catastrophically.
Analysis of the best practical manufacturing tolerances specifically related to the design of the components of a compressor as described in GB2215403 indicates that a working clearance of about 0.2 mm minimum would generally be required. If the clearance of the rotor at the apices possessed this higher value compared to the proposed 0.1 mm, and no apex seals were fitted as described in GB2215403, then the gas leakage at the apices would be undesirably high. Hence a design of compressor as described in GB2215403 has several deficiencies and would not result in the creation of an efficient machine.
Such deficiencies as these are no doubt the reason that the design of GB2215403, or any other design of epitrochoidal type machine, has failed to be successfully marketed for the general industrial compressor application despite it now being 60 years since the Wankel principles were first announced.
The only known production machine has been a small automotive air conditioning 2:3 type compressor manufactured for a time in the 1980s, as described in SAE 820159, U.S. Pat. No. 4,150,926, and “The Engineer” on 15/2/1979. This machine employed conventional apex seals. It suffered from a DV of 16%, a low volumetric efficiency and low energy efficiency.